[EMNLP'25] A Benchmark for Assessing VLM Safety with Real-World Memes
DongGeon Lee
oneonlee
AI & ML interests
Data-centric natural language processing, AI Safety
Recent Activity
upvoted
a
paper
about 1 month ago
What Users Leave Unsaid: Under-Specified Queries Limit Vision-Language Models
upvoted
a
collection
about 2 months ago
COMPASS
authored
a paper
about 2 months ago
Everyday Physics in Korean Contexts: A Culturally Grounded Physical
Reasoning Benchmark
Organizations
models
14
oneonlee/llama-3.1-nemoguard-8b-content-safety-merged
8B
•
Updated
•
1.83k
oneonlee/KorMath-Llama-3.1-8B-SFT
Text Generation
•
Updated
oneonlee/Llama-3.1-8B-SFT
Text Generation
•
8B
•
Updated
oneonlee/KorMath-Llama-3.1-8B
Text Generation
•
8B
•
Updated
oneonlee/Unverified-KorMath-Llama-3.2-1B-SFT
1B
•
Updated
oneonlee/Unverified-KorMath-Llama-3.2-1B
Text Generation
•
1B
•
Updated
oneonlee/KorMath-Llama-3.2-3B-SFT
Text Generation
•
3B
•
Updated
oneonlee/KorMath-Llama-3.2-1B-SFT
Text Generation
•
1B
•
Updated
oneonlee/Llama-3.2-1B-SFT
Text Generation
•
1B
•
Updated
oneonlee/KorMath-Llama-3.2-3B
Text Generation
•
3B
•
Updated
datasets
7
oneonlee/SIUO
Viewer
•
Updated
•
334
•
22
oneonlee/Meme-Safety-Bench
Viewer
•
Updated
•
50.8k
•
98
•
5
oneonlee/aegis2-train
Viewer
•
Updated
•
30k
•
11
oneonlee/wildguardmix-train
Viewer
•
Updated
•
86.8k
•
8
oneonlee/Wiki-NFQA
Viewer
•
Updated
•
1.89k
•
79
•
1
oneonlee/wiki-kg-prob
Viewer
•
Updated
•
230
•
4
oneonlee/cleansed_emocontext
Viewer
•
Updated
•
38.4k
•
20
•
2